Li, Lin Tzy; Yang, Seungwon; Kavanaugh, Andrea L.; Fox, Edward A.; Sheetz, Steven D.; Shoemaker, Donald J. (2011-06-01)
This poster presents one of our efforts developed in the context of Crisis, Tragedy, and Recovery Network (CTRnet) project. One of our derived works from this project is the use of social media by government to respond to emergency events in towns and counties. Monitoring social media information for unusual behavior can help identify these events once we can characterize their patterns. As an example, we analyzed the campus shooting occurred in the University of Texas, Austin, on September 28, 2010. In order to study the pattern of communication and the information communicated using social media on that day, we collected publicly available data from Twitter. Collected tweets were analyzed and visualized using Natural Language Toolkit, word clouds, and graphs. They showed how news and posts related to this event swamped the discussions of other issues.